Introduction
Of the estimated 2010 number of 7 · 7 million deaths worldwide in children under 5 years of age
, 49.6% occurred in sub Saharan Africa, 33% in South Asia and less than one percent in high income countries [
1]. Out of the estimated 8.8 million deaths occurring globally in the same age group in 2008, 41% occurred in neonates. Twenty-nine and 54% of these neonatal deaths were in Africa and South East Asia, respectively [
2]. South-Asian and sub-Saharan African regions account for two thirds of the global burden of neonatal deaths annually [
3].
Countries worldwide are striving towards achieving Millennium Development Goal 4, which deals with the reduction of neonatal deaths. High-income countries have made much progress while low-income countries still lag behind [
4]. High-income countries have managed to achieve an almost universal skilled birth attendance with provision of appropriate care. Neonatal mortality rate of as low as 0.45 per 1000 live births and intrapartum stillbirth rate of 1.22 per 1000 births have been observed [
4]. Countries with a neonatal mortality rate of greater than 30 per 1000 live births have about 50% skilled health worker attendance at birth [
5]. Skilled birth attendant at delivery, timely emergency obstetric care, provision of immediate newborn care and postnatal care are essential in promoting neonatal health [
4,
6,
7]. Therefore, a shift in place of delivery from home to health facilities is seen as an important strategy for improving neonatal outcomes [
8].
However, it is not only availability of health facilities for deliveries that is important, but also the quality of care provided. Emphasis is being placed on the quality of care, not only on availability of services [
9,
10]. Lack of quality care at health facilities limits women’s access to quality care [
11]. Women may deliver in health facilities, but still have poor perinatal and neonatal outcomes because of the substandard quality of care. A study in rural Tanzania showed that even at higher-level facilities where well trained health workers were supposed to be available, women experienced delays in receiving emergency obstetric care and had poor quality of care. Consequently, women experienced severe birth injuries and stillbirths [
12]. When women have a choice, they will go to health facilities where they perceive better quality of care, regardless of distance [
13,
14]. Forty-four percent of women by-passed their nearest health facility largely because of quality of care and delivered in another health facility [
15]. Women’s actual experience of care is significant and will greatly influence how women perceive quality. According to Hulton, Matthews and Stones (2000), a quality of care framework reflects both the provision of care and women’s actual experience of the care [
16]. It is argued that understanding women’s experiences of care is critical as it contributes to the use of health services and perinatal outcomes [
17].
Quality is not the only reason hindering pregnant women to access skilled birth attendance. Women continue to experience various problems to deliver with the help of skilled attendants. Literature suggests that women encounter sociocultural factors, perceived benefits, economic accessibility and physical accessibility as barriers to accessing skilled attendance during delivery [
18]. Women in sub-Saharan Africa still face limited access to skilled delivery, especially in the rural areas [
19,
20].
Malawi is making efforts to reduce intrapartum related deaths as a way of achieving Millennium Development Goal 4. This is, among other health interventions, asking pregnant women to deliver in health facilities with skilled attendance. In Malawi, the perinatal mortality rate is estimated at 40 deaths per 1,000 births and the neonatal mortality rate is 31 per 1,000 live births, with 71% skilled attendance for deliveries. Skilled attendance is higher in urban areas at 84% compared to 69% in the rural areas [
21]. In spite of these indicators, there is still need to increase the number of women delivering in health facilities as one way of preventing avoidable intrapartum-related deaths and neonatal deaths. This could further reduce the perinatal and neonatal mortality rates in Malawi. The argument here is that there is no justification for deaths due to intrapartum-related complications.
Despite the policy of Malawi government that women should deliver in health facilities with skilled attendants, it is not guaranteed that women will adhere and deliver in health facilities. Perception of quality is important in influencing the place of delivery, although it may not be the most important reason why women fail to access health facilities. Information on the views of women on perinatal care is limited in Malawi. The objective of this study was to explore the reasons why women delivered at home without skilled attendance despite receiving antenatal care at a health centre and their perceptions of perinatal care.
Methods
Study design
The study was a descriptive qualitative study that used a semi-structured interview guide. Face-to-face in-depth interviews were conducted. Participants who delivered outside a health facility between December 2010 and March 2011 were asked to describe their perception of perinatal care. The women were asked how they perceived the care they received during antepartum, intrapartum and postpartum. They were also asked about the information they received during provision of care. The interviews were conducted in the local language, Chichewa, using semi structured interview guide. During the interviews, follow-up questions using probes were asked in order to acquire a deeper understanding when an explanation was unclear. The interviews lasted on average, 45 minutes. All interviews were recorded, translated and transcribed verbatim in English.
Setting
The study was conducted in the Southern region of Malawi in the Chiradzulu district. Out of the 10 health centers under Chiradzulu district hospital, the catchment area of Namadzi health centre was selected for this study because many women from that area delivered at home with traditional birth attendants (TBAs) as revealed by hospital records of the period July 2008 to June 2009. The records showed that the Namadzi catchment area had the third largest number of expected pregnancies of 1380, but low skilled deliveries of 853 representing 62%. This was compared to the first two health centers that had 92% and 97% of skilled delivery respectively. Data were collected from mothers who received antenatal care at Namadzi health centre, but delivered in the community without skilled attendants.
Participants’ recruitment and data collection
Mothers who attended antenatal care at Namadzi health centre but delivered outside the health facility, either with TBAs or at home in the course of the study were selected using purposive sampling. All women who were approached and asked to participate in the study accepted. The age range covered the youngest primipara and the oldest delivering mothers in villages. This was done through reviewing postpartum records of mothers who came to the health centre after delivery as well as asking these mothers if they knew anybody else who had delivered in the community. The health surveillance assistants reviewed the records, and then went to the villages to trace the mothers. Health Surveillance Assistants (HSAs) were used in the villages to help trace the names of mothers that delivered during the study period at home and with TBAs from community members. The register of the Village Headmen did not have this information. The same HSAs traced the mothers in their homes. The first author provided information to and recruited potential participants that met the defined criteria. The HSAs organized with local leaders such as Village Headmen, headmasters and church leaders to provide a suitable place where the interviews were conducted. The first author conducted all the interviews. The interviews were done in a church, school building or out in the open air. The researcher ascertained that the place used for the interviews was appropriate to maintain privacy and comfort. A total of 12 mothers had been interviewed in the community when saturation of data was reached. There was no new information that was coming. Instead, there was repetition and confirmation of already collected data [
22].
The 12 mothers were from eight different villages namely Masuku, Walala, Kachere, Ulaya, Chelewani, Matola, Mng’omba and Lidala. Masuku village, 3.5 km from Namadzi health centre, is the closest one to the health centre. It is mostly flat land and it is close to the main road. Chelewani is a hilly area which is 4 km from the health centre. Lidala is flat with rivers and it is 5 km from the health centre. Ulaya is also 5 km from the health centre and it is located along the main road and it is flat. Kachere is located 6 km from the health centre and its terrain is hilly. Walala is 6 km from the health centre, along the main road. Matola is a flat area 7 km from the health centre. The furthest is Mng’omba, with hilly land and 10 km from the health centre. Its road is in bad condition with plenty of potholes. There is no public transportation available within the villages and only dry weather roads. The people have to walk to the main road in order to access transport such as minibuses to the health facility.
Data analysis
The Atlas. ti version 6.2 computer software program was used to code the transcripts and store the data. The Statistical Package for Social sciences (SPSS) version 18.0 was used for data entry and descriptive analysis of the participants. Analysis focused on participants’ perception of the care they received during antepartum, intrapartum and postpartum. Specifically, attention was on what they liked about the care they received and what problems or constraints they faced. All recorded interviews were transcribed verbatim in full. The analysis focused on developing coding categories where narrative information was organized according to emerging themes using thematic analysis [
23]. Coding of the data was done without fitting it into a preexisting coding frame. The data were read several times to identify themes that were related to quality of care.
Ethical consideration
Approval for the study was granted by the Norway Regional Committee for Medical Research Ethics as well as the College of Medicine Research and Ethics Committee (COMREC) in Malawi. A written permission to conduct the study was obtained from the District Health Officer (DHO) of Chiradzulu District Hospital.
Participants who agreed to participate in the study gave written informed consent or a thumbprint for those who were illiterate. Participation in the study was voluntary and participants were assured that anonymity would be observed at all times. Confidentiality of participants was maintained by using numbers on both the recorded interviews and transcripts.
Limitations
The study might suffer a bias because the participants were women who delivered at home within three months of the study. Consequently, it excluded the rest of the women who delivered at home who might have provided another view. Data were collected during the rainy season, possibly affecting movement of mothers to the health facility for delivery. It was not possible to go back to the community for member checking because of resource constraints. However, the findings are consistent with findings from other studies reviewed. The first author who collected the data is a health professional, which may have influenced some of the participants’ responses.
Discussion
A way of reducing perinatal morbidity and mortality is to get more pregnant women to deliver with the assistance of skilled attendants. Understanding perceived barriers that prevent pregnant women from delivering in health facilities is a step towards focusing on how to help pregnant women to reduce or eliminate these factors.
Of interest on the demographic characteristics of the participants is that most women who delivered at home were those that were previously classified as low risk. This is in terms of their age and parity. Risk assessment was done antenatally and clients were explicitly recommended where to deliver [
18]. The low risk category of women was even allowed to deliver with trained TBAs before the concept of skilled birth attendants for all became the focus. It may be the same reasoning behind the finding that when the women knew during antenatal care that they had no problems or complications with the pregnancy, they felt it was safe to deliver outside a health facility. A study in rural Tanzania demonstrated that women with ‘normal’ pregnancies expected to deliver at home with no problems. Health workers reported antenatal care provided the women with reassurance that their pregnancies are normal [
24,
25].
Evidently there is need to increase women’s awareness on the necessity of skilled birth attendants during delivery to ensure that women deliver in health facilities and neonates are given appropriate care. Women need to understand through targeted health information messages that complications may occur without warning anytime during labor and delivery; and even after completely ‘normal’ pregnancies.
It is not possible to say why fewer than half of the participants who delivered outside the health facility were from Masuku village. This is closest to the health centre out of the rest of the villages. Its terrain does not make this village more difficult to access the main road than that of the rest of the villages. However, more than half of participants from this village talked about the issue of self-delivery in health facilities while some had beliefs that witchcraft would make pregnant woman delay in giving birth. An earlier study done in Malawi also found that more than half of the participants believed it was possible to be bewitched during labor. This belief did not change post intervention [
26]. Maimbolwa, Yamba, Diwan, & Ransjo-Arvidson (2003) showed that a pregnant woman should not reveal labor had began to avoid complications during labor and delivery caused by evil spirits and witches in Zambia [
27]. The women may prefer to wait and go to a health facility until labor is well established, so that it cannot to be stopped by witchcraft. In this way, they delay going to the health facility or they do not comply at all. Choudhury & Ahmed (2011) revealed that in rural Bangladesh, traditional beliefs delayed care-seeking in health facilities [
28]. This belief about witchcraft needs to be addressed for communities not to relate labor and delivery process to witchcraft. Of course, a simpler mechanism might be that those who are closest to the clinic take the risk of waiting until the last minute, and then most often fail to reach there. Provision of appropriate perinatal information by health workers is necessary to raise community awareness.
Additionally, health workers must strive to avoid self-deliveries when women go to deliver in health facilities. Self-deliveries deter women using health facilities with subsequent pregnancies [
29,
30]. Use of a companion to support a woman during labor is associated with a safe and satisfying birth experience [
31]. Due to shortage of health workers they are not in a position to meet all the needs of a woman in labor. Guidelines in Malawi for the provision of care stipulate that a support person be present during labour and delivery [
32,
33]. Using a doula to be with the women provides a one to one support to the women as they do not take care of anyone else [
34,
35]. A study in Malawi has shown that supportive companion during childbirth is highly acceptable among mothers, health professionals, and the community [
36]. Implementation of this strategy may address women’s concern of being left alone during labor in health facilities in Malawi.
Women’s views on quality focused on how they were received at antenatal clinic and when they went to the health facility after delivery. Participants stated care was good at antenatal clinic and in labor ward based on good reception and what was done. The women did not complain about the technical quality of care. This is similar to findings of women who delivered at a district hospital in the same district [
37]. The authors have shown that women are often not critical to the care they receive. The women did not know the quality of care to expect because they were not well informed, ending with higher risk for delivery problems. The major concern for the women in this study was poor staff attitudes. Health workers shouted at the women and even threatened to beat the women if they would be troublesome when they went for delivery. Other studies confirm that poor attitudes from health workers, who were rude and abused women, discouraged the women from delivering in health facilities [
19,
24,
30,
38,
39]. This may provide an explanation why some of the women did not deliver at the facility. It was perhaps fear of abuse from health workers though they said otherwise. Humane aspect of care matters greatly in provision of care to women during labor and delivery [
40,
41]. A laboring woman is vulnerable and in pain thus needs understanding. Positive attitudes and empathy of health workers was related to delivery in health facilities [
39,
40,
42].
There is an urgent need to address the poor attitudes of health workers for them to provide appropriate professional midwifery care to women. Frequent supportive supervision of health workers at the health facility is necessary to resolve problems they experience that negatively impact on provision of care. This is important, as health centers are understaffed; the numbers are not adequate against the workload affecting provision of care [
43,
44]. Management support and fairness in managerial practices contribute to improving health workers’ motivation and performance [
45‐
47] resulting in provision of optimal care to clients.
Delivery at a TBA was associated with inadequate care because the TBA is not in a position to identify and manage complications. Despite knowing this, women still ended up delivering at TBAs and home. Probably the women felt they would have normal deliveries therefore; there was no need to go to a health facility. Research findings in Indonesia revealed that some community members preferred TBAs and home deliveries despite presence of a village midwife in the village. Specifically, services of skilled birth attendants were perceived important only during complications [
48,
49]. Delivery process was viewed as easy by Bolivian women therefore there was no need for them to go to the hospital [
50]. Comparable, women’s experiences of uncomplicated home birth made these women think that delivery in a health facility was not necessary [
30].
Furthermore, the women in this study could not go to wait at the health facility because they had no guardians to take care of their needs whilst there and other responsibilities at home. Therefore, the women may have preferred to go in established labor, deliver, then go back home. In view of this, the solution by authorities that all pregnant women wait at health facilities for delivery is not feasible for these women. This is regardless of various factors; time of the day when labor starts, lack of people to escort a laboring woman at night; and rapid labor and rainy season that hindered the women’s access to health facility. The TBAs are then seen as a ready source of help when these problems that are beyond their control hinder access to health facilities [
19,
24]. Habit also played a significant role, like a grandmultipara who delivered only a first child in a health facility and subsequently delivered at TBAs with no problems. Therefore, mention of rapid labor or labor starting at night could be an alibi for home deliveries [
51]. Health workers should be aware of these problems and discuss them with the women during antenatal care and communities when they discuss birth preparedness. There is also need for a more thorough exploration of these barriers that prevent women from accessing health care for better understanding; and, subsequently work with the communities to identify best solutions that will be ideal for them.
As women continue to deliver with TBAs and at home, there is need to improve initial care given to babies by using health surveillance assistants that are already trained in community based maternal newborn care package (CBMNC) or train them in places where they are not available [
52]. This is an important strategy as, currently, Malawi is still far from providing skilled birth attendants at home.
Distance has been recognized as a major barrier to delivery in health facilities [
18,
20,
53,
54] but distance was not found to be an issue in this study.
Assessments of some mothers and babies or delayed assessments when they went to the health facility after the mothers had delivered was inappropriate. Babies are vulnerable especially during the first 24 hours after delivery [
55] and thorough assessment during this time is consequently not only necessary, but crucial. The babies merely had cord care and weighing, with few babies that did not have their birth weight checked. This may discourage other mothers who deliver outside health facilities to go to health facilities as required for review and appropriate management. Guidelines stipulate that women who deliver outside of a health facility should go to a health facility for a postnatal check-up within 48 hours of giving birth [
56]. Only 21% of babies born at home in Malawi had a postnatal check after delivery within 42 days and barely 18% within 48 hours of delivery. It was even less for Chiradzulu district at 6% and 10% within 24 and 48 hours of delivery, respectively [
57]. Essential neonatal care practices such as keeping the neonate warm, cord care and initiation of exclusive breastfeeding may not be done or wrongly done outside the health facility. Skilled attendants must provide care to a neonate immediately after delivery [
58]. It is therefore imperative that all the babies are assessed and given suitable care when they arrive at a health facility. A baby was wiped with warm water, cord care done then wrapped. This indicates some sort of assessment was done on the baby but the mother felt that the baby was not examined. This is probably because the mother was not told about the examination. Consequently, there is need for health workers to improve in the way they communicate with clients.
Appropriate information on baby care and danger signs of the baby should be provided to all mothers before discharge. This will enable mothers to properly take care of their babies and seek care promptly when they encounter any problems [
59,
60].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
LCK conceptualized the study, collected, analyzed data and drafted the manuscript. JØO is the Principal supervisor of the study. GB, EM, AM and JØO have critically reviewed drafts, edited and provided important intellectual content. All authors read and approved the final manuscript.